The response of the colon to aldosterone is believed to be an important adaptive mechanism to excessive sodium losses in diarrhea. However, the degree to which mineralocorticoid activity actually influences fecal output of sodium in people with diarrhea is unknown. To gain insight into this question, 10 normal people were treated with placebo, fludrocortisone (an aldosterone agonist), and spironolactone (an aldosterone antagonist) during three experimental periods lasting nine days. On days 5-8, diarrhea was induced by ingestion of phenolphthalein. Diet was controlled. Fecal sodium was 40 meq/day on placebo and 29 meq/day on fludrocortisone, consistent with mineralocorticoid stimulation of intestinal sodium absorption. However, contrary to our expectations, spironolactone therapy was also associated with a fall in fecal sodium output, to 28 meq/day. To explain this paradoxical effect of spironolactone, we suggest that sodium depletion caused by spironolactone's natriuretic action on the kidney caused the release of an unknown stimulant of intestinal sodium absorption, whose action more than overcame the reduced colonic absorption resulting from inhibition of aldosterone activity by spironolactone. This interpretation implies that the intestinal adaptation to sodium depletion in diarrhea involves both aldosterone and an aldosterone independent factor, working in concert to reduce fecal sodium output.